random thoughts and analyses that may or may not relate to mosquitoes and mosquito-borne diseases
i have no timeline OR expectation of how often i will maintain this section of my website
this information reflects my opinions only
this information is not peer-reviewed
August 19th, 2025
Please don't take away Google Scholar....!
July 7th, 2025
Figuring out how to write is a hallmark of being a scientist. Using AI/LLMs won't make you a better writer.
June 18th, 2025
The irony of posting this link is not lost on me (I am an editorial board member at Scientific Reports)
Also, the DAFNEE link seems like a promising resource (that I was unaware of) for myself, my lab members, and others in the field to find suitable and quality-focused journals for our research
June update: I was working on new ideas/ways to visualize measles outbreak data in our region, and then life caught up with me. Hopefully they've archived some of this data...
April 15th, 2025
I missed the non-existent weekly deadline, so we're a few days late as I play with numbers waiting for an HVAC mechanic...!
As of April 11th, there were 541 total reported measles cases in TX, so about 60 new cases (a 12.5% increase in case numbers from the following week) - I'm glad that number was lower than predicted, and it just goes to show that epidemics - in real-time - are dynamic. However, as of April 11th, 22 counties in Texas had reported at least one measles cases since the outbreak began - which was an increase in 3 counties from the prior week. Checking the DSHS update today, one more new county has reported a measles case (bringing the total number of counties that have reported at least one measles case linked to the on going outbreak to 23).
I'd like to commend the TX DSHS for drastically changing the way they report bi-weekly data on their website. Most of it is downloadable in some form, which makes it a little easier to make pictures with the data. Today, we take a look at trends in the daily reported measles cases in Texas.
While DSHS reports cases through April 9th, they mention that new cases could be found/reported going all the way back to March 26th (the date of infection seems to be the date the characteristic rash appears).
The top plot shows final daily case counts in black and conditional daily case counts in red (based on the rash date criteria). The blue line with the grey shaded region shows a non-linear fit to the confirmed data - just something to aid a viewers identification of an increasing trend in case reports.
The bottom plot shows the average number of daily cases reports per week - the points represent the average daily report and the whiskers represent the error when estimating the mean. I think this plot makes it a little easier to see that - overall - case reports have been increasing daily, and we're up to around 11 new cases reported each day in Texas.
Depending on the prediction method, the overall trend is that daily case reports are going to increase over the next few weeks.
The plot on the left shows the current reported cases as red dots (through March 25th) and the grey shaded regions shows the predicted range of cases based on a generalized linear model (GLM) into April 10th. While a GLM provides a simpler fit to the data (it assumes a linear relationship between variables), its not really capturing uncertainty moving from March 26th to April 10th (which is the 100th day of the year). Remember, cases have been reported up to April 9th, its just that final case counts are only confirmed through March 25th.
The plot on the right shows the same thing but with a more flexible additive model. This model assume the relationship between time and number of cases being reported in non-linear (e.g., it can take any shape). Overall, the GAM predicted somewhat stable daily case counts for about a month (sometime mid February to late March); the uptick in cases right at the end of March pushes predictions in to the future up. You can see how the GAM prediction starts widening as it moves into time periods unseen by the model - thus its showing confidence in new reported cases decreases moving forward. Let's hope the lower end of the prediction (the piece turning downward) is what we see over the next few weeks as case counts continue to be finalized.
April 6th, 2025
New weekly total case reports in Texas are increasing somewhere around an average of 33% each week. Rather than create some sort of figure that projects what the number of new cases could look like over the next few months, let's just focus on what that means moving forward to the end of next week. Using the average, that means you'd predict 159 new cases next week. More realistically, new cases in Texas have increased somewhere between 19% and 60% in the last five weeks. If measles transmission continues to spread within this range, next week there could be anywhere between 93 and 171 new reported cases in TX. This is just a prediction based on past observed changes in weekly case counts; while I expect numbers to increase next week, lets hope we start seeing overall declines in the rate of new cases.
School (ISD or Private) and county level vaccination rates among kindergarten and high school age children is available in downloadable form TX DSHS. This information is available as a searchable database, so to make it a bit more helpful at a state-level, I created a map using some of the school/ISD data. Community-level protection (e.g., herd immunity) against measles is estimated at somewhere above 92%, so I tallied up the number of schools reported in each county with vaccination rates above or below that minimum threshold**. The map on the left shows the percent of schools in each county (regardless of ISD or Private) that are below that threshold (blue = lowest value, which is good because that means there is a greater prevalence of schools with coverage at or above 92%; red - highest, which is bad because that means there is a greater prevalence of schools with coverage below 92%; white - all reported schools in the county meet or exceed the 92% threshold). The dots on the map represent the number of new reported cases last week. I think the main message is that the West Texas region as well as many other parts of TX are more vulnerable to continued measles transmission in the age groups most at risk of severe complications than you can infer from a list of schools online. The state also reports data at a high school level; I have not looked at that data at the moment.
10 new cases reported in Lubbock as of the April 4th report, and three new public exposure sites have been reported as of March 31st. The three new sites are not great in terms of mass exposure: an event/recreation center often frequented by young children, a shopping center, and a daycare. It occurred to me that these reported sites are just the places people went where the general public may also have visited; I wonder, how many friends/neighbors are people (un)knowingly visiting while infectious????
March 31st, 2025
Cases in four states have been linked to the initial outbreak in West TX. While this is consistently reported in the news, its difficult to get a feel for this visual. After collating information from different state websites/social media postings, the map in the left panel shows the counties reporting measles cases in Texas (400 cases), New Mexico (44 cases), Kansas (23 cases), and Oklahoma (9 cases). Kansas and Oklahoma are pretty much only reporting the location of cases separate from the total (I had to make a guess as to how many were positive in each county). Also, most states are just updating the bi-weekly totals of cases, so unless you're paying attention and writing things down, its difficult to tell how many new cases are occurring each week - this is what the right panel is showing. I could only figure this out for TX and NM, since that is that I could find information on. Overall, cases continue to increase every week AND new cases are being reported in new places each week (roughly 3 new counties tested positive in TX last week!)
13 new cases were reported in Lubbock last week, and the city has published a list of locations where mass exposure events could have occurred. It's hard to get an idea of how those sites pan out on a map, so I geo-coded the provided addresses to see if there was any aggregation in the locations (e.g., maybe people from outside Lubbock are heading to things like Costco/Walmart/HEB which are on the edge of town). Welp, exposure is pretty universal throughout the City. As you can see from the map, there really isn't any indication of a spatial pattern worth exploring statistically. Take home message is that exposure can happen just about anywhere as people continue to live their lives pre/post known/unknown measles exposure events.
I'm still working out the best way to create/present predictions into the future, but the plots on the left outline some projections going forward into early summer. The predictions assume geometric growth rates, and given that all three states continue to report new cases weekly, case numbers are expected to increase into the foreseeable future. Please keep in mind that the y-axis in these plots is on a log10 scale, so 1 = 10, 2 = 100, 3 = 1000, and 4 = 10000. Also, red points indicate reported case numbers and the grey lines are the simulated predictions.
March 23rd, 2025
Two measles cases were reported in Gaines County, TX in late January of this year. Since then, over 300 cases have been reported in TX alone. I have been following the news releases that TX DSHS has been releasing on Tuesdays and Fridays of each week. Overall, the outbreak has received national attention, though I have had a hard time trying to understand trends in the data. So, I got side-tracked this weekend trying to create different figures and maps that break down trends in the data.
I'll note here that there are other things in the data such as general age breakdown (0-4 yo, 5-18 yo, 18yo+) and vaccination status - at the moment, I do not intend to look at those data, and the plots/maps below only contain information on case totals and counties reporting cases.
Weekly updates provide only the total number of cases detected since the initial index case. From those updates, its hard to tell how rapidly the outbreak is spreading in terms of number of cases per week.
The above panel plots shows three things: The left plot shows the number of reported cases in TX from each update (since 1/30/25) as grey circles. I then made a few predictions going forward the next three weeks (yellow, orange, red triangles). To make these predictions, I calculated the geometric growth rate in cases (e.g., the percent change in cases from one report to the next). Then, using the total number of reported cases in TX on 3/21/25, I projected forward using the observed minimum (yellow), average ( orange), and maximum (red) growth rate in the data set. To generate measures of uncertainty, I did two simulations. The first simulation (middle panel plot) projects 1000 different simulations through the next few months by taking random draws from the entire distribution of observed growth rates (7% - 25% weekly increase). The second simulation (right panel plot) projects 1000 different simulations through the next few months by taking random draws around the average observed growth rate (mean 10% weekly increase, standard deviation 6%). Many MANY caveats and assumptions are occurring here and making epidemic predictions can be a fraught endeavor. I think if anything I'm trying to give a quantitative perspective & a visual that reinforces the context/sentiment in the news that controlling the outbreak is incredibly difficult.
The maps provided by the various health agencies involved in the outbreak response seem to only list the location of detected cases. The maps provided here give greater context to that information. The smaller plots show presence (Red), absence (blue) of cases in the 14 counties in TX that have reported a measles case linked to the Gaines county cases. The bigger plot with TX and NM shows how many weeks a county has reported a case since the index cases.
From the initial index case, 16 counties between TX and NM have reported measles cases related to the index case. As of last week (3/17 - 3/22), three additional counties in TX are reporting cases; three counties reported their first infections two weeks ago (3/10 - 3/14).
Overall, each week an additional two counties are reporting measles cases. This isn't a real spatial prediction (that's more complicated) - its simply here to again provide some context to the spread of the outbreak. Possibly more to come on this in the future.
